Act 3: Lying About Storage

How does FAT32 link file fragments across data blocks?

Linked List Structure: FAT32 uses a centralized File Allocation Table (FAT) where each 32-bit entry corresponds to a physical data block. The value stored inside an entry points to the block index containing the next segment of the file, forming an on-disk linked list.

Given a FAT32 partition using $4\text{ KB}$ clusters, calculate the size of one File Allocation Table (FAT) if there are $2^{28}$ addressable blocks.

FAT Size Calculation:

$$\text{FAT Size} = 2^{28} \text{ entries} \times 4 \text{ bytes/entry} = 2^{30} \text{ bytes} = 1\text{ GB} \text{ per FAT table} \text{[cite: 78, 80]}$$

What is the fundamental factor limiting the maximum single file size on a FAT32 volume to $4\text{ GB}$?

4 GB File Size Limit:

• The directory table entry in FAT32 uses a 32-bit field to record file size in bytes.

• $2^{32} \text{ bytes} = 4,294,967,296 \text{ bytes} = 4\text{ GB}$.

What information is stored inside an Ext4 Inode vs. a Directory Table?

• Inode: Stores all metadata for a specific entity (file permissions, owner UID, file size, timestamps, extent tree root describing physical data block locations). It does not store the filename.

• Directory Table: Acts as a special file mapping human-readable filenames directly to global Inode numbers

Explain how Ext4 extents replace block pointer arrays to increase storage contiguity and lookup efficiency.

• An extent represents a single contiguous run of physical blocks using a tuple: [Logical Block Start, Length, Physical Block Start].

• Instead of storing individual block pointers for every $4\text{ KB}$ block, a single extent can map up to $128\text{ MB}$ of contiguous data, keeping lookups fast and reducing metadata storage overhead.

Trace how Ext4 handles file fragmentation when a file requires more than 4 extents.

An inode has space for 4 extents. When a 5th extent is needed, EXT4 can use an extent tree to store and organise the additional extent information. It does not mean the file can only have 4 extents.

Define the Crash Consistency Problem in storage systems.

Crash Consistency Problem: The risk that a sudden power failure or kernel crash during a multi-step filesystem operation (e.g., updating bitmaps, writing inodes, writing data) leaves on-disk data structures corrupted or in an inconsistent state.

Detail the exact four-step process Ext4 uses to perform a journaled write.

Ext4 Journaling Steps:

1. Data Write: Write payload data to allocated data blocks and flush to disk.

2. Journal Metadata Write: Write all impending metadata changes (inodes, bitmaps, directory entries) into the circular journal file and flush to disk.

3. Journal Commit: Write a commit record containing a transaction sequence number and checksum to the journal and flush.

4. Checkpointing: Copy the metadata changes from the journal to their permanent physical locations on the filesystem.

Why does FAT32 lack crash consistency compared to Ext4?

FAT32 Limitations: FAT32 writes metadata directly to disk without a transaction log or journal. If a crash occurs midway through updating duplicate FAT tables or directory entries, the filesystem structures become permanently desynchronized, leading to lost clusters or broken file chains.